Remove Regularization section.

I can't find values of regularization strength that matter for this problem.

Author: MarkDaoust

samples/core/tutorials/estimators/wide.ipynb

@@ -6,10 +6,10 @@
       "name": "wide.ipynb",
       "version": "0.3.2",
       "provenance": [],
-      "toc_visible": true,
       "collapsed_sections": [
         "MWW1TyjaecRh"
-      ]
+      ],
+      "toc_visible": true
     },
     "kernelspec": {
       "display_name": "Python 3",
@@ -33,7 +33,6 @@
         "id": "mOtR1FzCef-u",
         "colab_type": "code",
         "colab": {}
-        
       },
       "cell_type": "code",
       "source": [
@@ -1077,9 +1076,8 @@
         "        [age_buckets, 'education', 'occupation'], hash_bucket_size=1000),\n",
         "]\n",
         "\n",
-        "model_dir = tempfile.mkdtemp()\n",
         "model = tf.estimator.LinearClassifier(\n",
-        "    model_dir=model_dir, feature_columns=base_columns + crossed_columns)"
+        "    model_dir=tempfile.mkdtemp(), feature_columns=base_columns + crossed_columns)"
       ],
       "execution_count": 0,
       "outputs": []
@@ -1138,8 +1136,8 @@
       "source": [
         "results = model.evaluate(test_inpf)\n",
         "clear_output()\n",
-        "for key in sorted(results):\n",
-        "  print('%s: %0.2f' % (key, results[key]))"
+        "for key,value in sorted(result.items()):\n",
+        "  print('%s: %0.2f' % (key, value))"
       ],
       "execution_count": 0,
       "outputs": []
@@ -1199,67 +1197,7 @@
       "source": [
         "If you'd like to see a working end-to-end example, you can download our\n",
         "[example code](https://github.com/tensorflow/models/tree/master/official/wide_deep/census_main.py)\n",
-        "and set the `model_type` flag to `wide`.\n",
-        "\n",
-        "## Adding Regularization to Prevent Overfitting\n",
-        "\n",
-        "Regularization is a technique used to avoid **overfitting**. Overfitting happens\n",
-        "when your model does well on the data it is trained on, but worse on test data\n",
-        "that the model has not seen before, such as live traffic. Overfitting generally\n",
-        "occurs when a model is excessively complex, such as having too many parameters\n",
-        "relative to the number of observed training data. Regularization allows for you\n",
-        "to control your model's complexity and makes the model more generalizable to\n",
-        "unseen data.\n",
-        "\n",
-        "In the Linear Model library, you can add L1 and L2 regularizations to the model\n",
-        "as:"
-      ]
-    },
-    {
-      "metadata": {
-        "id": "cVv2HsqocYxO",
-        "colab_type": "code",
-        "colab": {}
-      },
-      "cell_type": "code",
-      "source": [
-        "#TODO(markdaoust): is the regularization strength here not working?\n",
-        "model = tf.estimator.LinearClassifier(\n",
-        "    model_dir=model_dir, feature_columns=base_columns + crossed_columns,\n",
-        "    optimizer=tf.train.FtrlOptimizer(\n",
-        "        learning_rate=0.1,\n",
-        "        l1_regularization_strength=0.1,\n",
-        "        l2_regularization_strength=0.1))\n",
-        "\n",
-        "model.train(train_inpf)\n",
-        "\n",
-        "results = model.evaluate(test_inpf)\n",
-        "clear_output()\n",
-        "for key in sorted(results):\n",
-        "  print('%s: %0.2f' % (key, results[key]))"
-      ],
-      "execution_count": 0,
-      "outputs": []
-    },
-    {
-      "metadata": {
-        "id": "5AqvPEQwcYxU",
-        "colab_type": "text"
-      },
-      "cell_type": "markdown",
-      "source": [
-        "One important difference between L1 and L2 regularization is that L1\n",
-        "regularization tends to make model weights stay at zero, creating sparser\n",
-        "models, whereas L2 regularization also tries to make the model weights closer to\n",
-        "zero but not necessarily zero. Therefore, if you increase the strength of L1\n",
-        "regularization, you will have a smaller model size because many of the model\n",
-        "weights will be zero. This is often desirable when the feature space is very\n",
-        "large but sparse, and when there are resource constraints that prevent you from\n",
-        "serving a model that is too large.\n",
-        "\n",
-        "In practice, you should try various combinations of L1, L2 regularization\n",
-        "strengths and find the best parameters that best control overfitting and give\n",
-        "you a desirable model size."
+        "and set the `model_type` flag to `wide`."
       ]
     },
     {
@@ -1332,4 +1270,4 @@
       ]
     }
   ]
-}
+}